labyrinthinefandomcom-20200215-history
"Computing Machinery and Intelligence"
Alan Turing 1912-1954 Turing was a British mathematician, logician, cryptanalyst and computer scientist. His worked helped to develop computer science as a "field" and formalized the concept of the algorithm and computation with his theoretical Turing machine, which is considered a conceptual precursor to the modern computer. In WWII, Turing worked a codebreaker at Bletchley Park, and cracked the legendary Enigma machine. In 1952 Turing was prosecuted for homosexuality and accepted chemical castration rather than go to prison. He died in 1954 from cyanide inserted in an apple; whether suicide, accidental, or purposeful is unclear. On September 10, 2009, then-British Prime Minister Gordon Brown made an official public apology on behalf of the British government for the way in which Turing was treated after the war. Probably the most enduring legacy of Turing's research has been the "Turing test", shorthand for a test which demonstrates computer intelligence. Admittedly, the imitation games Turing cites are not clearly explained, and thus it is ambiguous as to whether Turing intended for the test to prove whether a machine can imitate a human, or make someone believe it is human. Computing Machinery and Intelligence 1950 Turing essay was published in MIND, a philosophical journal, not directed toward computer scientists (this is also demonstrated by its heavy drawing on refutations of the opposition to computer intelligence). This essay is center in the major philosophical issues that divide thought on the nature of the brain--essentially dualist (the brain has some non-physical properties and cannot be composed in physical forms) or materialist (the brain is merely an object that can be studied fully). Turing opens with the question of "Can Machines Think" but deposes it in favor of describing the problem in terms of a game. Turing offers a standard variation of the imitation game, in which subject C is trying to guess which sex subject A and B are. A is male and trying to dissuade, and B is female and trying to convince. C may ask A and B any questions and they must answer about themselves truthfully. Turing then consider if C would decide wrongly as often if A were played by a computer (a computer programmed to behave as a male trying to decieve C about his appearance), as C would when A and B are people. This forms, initially, the new question for Turing which can be sumarised something like: "Let us fix our attention on one particular digital computer C. Is it true that by modifying this computer to have an adequate storage, suitably increasing its speed of action, and providing it with an appropriate programme, C can be made to play satisfactorily the part of A in the imitation game, the part of B being taken by a man?" The question is also draws "a fairly sharp line between the physical and the intellectual capacities of a man" -- thus we are not asking a computer to appear as human. The computer Turing envisions is hypothetical--the idea of a digital computer is that it can carry out any operation which could be done by a human computer, following fixed rules: "We are not asking whether all digital computers would do well in the game nor whether the computers at present available would do well, but whether there are imaginable computers which would do well" (4). ''' Components of a Digital Computer Turing defines the digital computer as having 3 components: store, executive unit, control. '''Store is a store of information--the rules, the paper, for a human even part of the memory. Executive Unit '''is the part which carries out the various individual operations. '''Control '''makes sure these instructions are obeyed correctly and in the right order. '''Programming '''is "contructing instruction tables" (6). Digital computers are universal machines'. They can mimic any discrete state machine, and any calculation can be done with any digital computer, assuming it is properly programmed and have enough store. In this sense, all digital computers are equivalent. Philosophical Opposition to Computer Intelligence Turing counters major contrary views: Theological Objection: thinking is function of man's immortal soul and only man has been given the gift of thinking by God "Heads in the Sand" Objection: the consequences of computer thoughts are too awful to image (Weizenbaum's objection perhaps?) The Mathematical Objection: since computers answer questions based on logic, there is a limit to what they can compute Argument from Consciousness: thinking involves emotions; Turing counters by saying there is no way to know if other's experience emotions Argument from Various Disabilities: computers can't do "X" --be self-aware, make mistakes, etc. Lady Lovelace's Objection: machines are incapable of independent learning Argument from Continuity in the Nervous System: the nervous system is not digital Argument from Informality of Behavior: any system governed by laws cannot be intelligenct Argument from Extrasensory Perception: a computer could not have its mind read through ESP The Learning Machine Turing suggests that rather than building a computer with an adult learning system, we should build a computer like a child--that can learn. Turing articulates that 3 things bring a mind to the state it is in: initial state of the mind ('birth'), '''education '''it has received, and other '''experiences'. Thus, for a child, all you need is a child program and an education process. Turing covers how it would be easy to program a computer to learn that it would repeat approved behaviors and not repeat disapproved behaviors. This idea that the mind is a feedback relay shows up again in Weiner. For Turing, the computer is the hardware, the child program is the software (and so to would be the education process?).